| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 57385, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 9.945543260433913e-06, | |
| "loss": 3.7697, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 9.891086520867823e-06, | |
| "loss": 2.9544, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 9.836629781301735e-06, | |
| "loss": 2.4309, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 9.782173041735646e-06, | |
| "loss": 2.1416, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 9.727716302169558e-06, | |
| "loss": 2.0346, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 9.673259562603468e-06, | |
| "loss": 1.9859, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 9.61880282303738e-06, | |
| "loss": 1.9512, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 9.56434608347129e-06, | |
| "loss": 1.9171, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 9.509889343905202e-06, | |
| "loss": 1.8989, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 9.455432604339114e-06, | |
| "loss": 1.868, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 9.400975864773026e-06, | |
| "loss": 1.8423, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 9.346519125206936e-06, | |
| "loss": 1.8311, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 9.292062385640848e-06, | |
| "loss": 1.8139, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 9.237605646074758e-06, | |
| "loss": 1.809, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 9.18314890650867e-06, | |
| "loss": 1.7932, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 9.128692166942582e-06, | |
| "loss": 1.7807, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 9.074235427376494e-06, | |
| "loss": 1.7729, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 9.019778687810404e-06, | |
| "loss": 1.7695, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 8.965321948244316e-06, | |
| "loss": 1.7464, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 8.910865208678226e-06, | |
| "loss": 1.7436, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 8.856408469112138e-06, | |
| "loss": 1.736, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 8.80195172954605e-06, | |
| "loss": 1.729, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.7001079320907593, | |
| "eval_runtime": 3.4215, | |
| "eval_samples_per_second": 39.748, | |
| "eval_steps_per_second": 4.969, | |
| "step": 11477 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 8.747494989979961e-06, | |
| "loss": 1.7264, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "learning_rate": 8.693038250413871e-06, | |
| "loss": 1.708, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 8.638581510847783e-06, | |
| "loss": 1.6948, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "learning_rate": 8.584124771281695e-06, | |
| "loss": 1.698, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "learning_rate": 8.529668031715605e-06, | |
| "loss": 1.692, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 8.475211292149517e-06, | |
| "loss": 1.684, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 8.420754552583429e-06, | |
| "loss": 1.6879, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "learning_rate": 8.36629781301734e-06, | |
| "loss": 1.6804, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "learning_rate": 8.311841073451251e-06, | |
| "loss": 1.6713, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 8.257384333885163e-06, | |
| "loss": 1.6703, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 8.202927594319073e-06, | |
| "loss": 1.6646, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "learning_rate": 8.148470854752985e-06, | |
| "loss": 1.651, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "learning_rate": 8.094014115186897e-06, | |
| "loss": 1.6488, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "learning_rate": 8.039557375620808e-06, | |
| "loss": 1.6452, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "learning_rate": 7.985100636054719e-06, | |
| "loss": 1.6386, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "learning_rate": 7.93064389648863e-06, | |
| "loss": 1.6349, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 7.87618715692254e-06, | |
| "loss": 1.6345, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "learning_rate": 7.821730417356452e-06, | |
| "loss": 1.6294, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "learning_rate": 7.767273677790364e-06, | |
| "loss": 1.631, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 7.712816938224276e-06, | |
| "loss": 1.6261, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 7.658360198658186e-06, | |
| "loss": 1.6281, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "learning_rate": 7.603903459092098e-06, | |
| "loss": 1.611, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "learning_rate": 7.549446719526009e-06, | |
| "loss": 1.6155, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.596663475036621, | |
| "eval_runtime": 3.4296, | |
| "eval_samples_per_second": 39.655, | |
| "eval_steps_per_second": 4.957, | |
| "step": 22954 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 7.49498997995992e-06, | |
| "loss": 1.6029, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "learning_rate": 7.440533240393831e-06, | |
| "loss": 1.607, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "learning_rate": 7.386076500827744e-06, | |
| "loss": 1.5977, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "learning_rate": 7.331619761261655e-06, | |
| "loss": 1.5922, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "learning_rate": 7.277163021695566e-06, | |
| "loss": 1.5956, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "learning_rate": 7.222706282129477e-06, | |
| "loss": 1.5855, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "learning_rate": 7.168249542563388e-06, | |
| "loss": 1.5826, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "learning_rate": 7.1137928029972995e-06, | |
| "loss": 1.5846, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "learning_rate": 7.059336063431211e-06, | |
| "loss": 1.5899, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 7.004879323865122e-06, | |
| "loss": 1.5828, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "learning_rate": 6.950422584299033e-06, | |
| "loss": 1.5762, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "learning_rate": 6.895965844732945e-06, | |
| "loss": 1.5739, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "learning_rate": 6.841509105166856e-06, | |
| "loss": 1.574, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "learning_rate": 6.787052365600767e-06, | |
| "loss": 1.5759, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "learning_rate": 6.732595626034678e-06, | |
| "loss": 1.5737, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "learning_rate": 6.67813888646859e-06, | |
| "loss": 1.5637, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "learning_rate": 6.623682146902502e-06, | |
| "loss": 1.5635, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "learning_rate": 6.569225407336413e-06, | |
| "loss": 1.5641, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "learning_rate": 6.514768667770324e-06, | |
| "loss": 1.553, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "learning_rate": 6.460311928204235e-06, | |
| "loss": 1.5699, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "learning_rate": 6.405855188638146e-06, | |
| "loss": 1.5695, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "learning_rate": 6.3513984490720584e-06, | |
| "loss": 1.5665, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "learning_rate": 6.296941709505969e-06, | |
| "loss": 1.5527, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 1.5436657667160034, | |
| "eval_runtime": 3.2624, | |
| "eval_samples_per_second": 41.687, | |
| "eval_steps_per_second": 5.211, | |
| "step": 34431 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "learning_rate": 6.24248496993988e-06, | |
| "loss": 1.5506, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "learning_rate": 6.188028230373791e-06, | |
| "loss": 1.559, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 3.09, | |
| "learning_rate": 6.133571490807702e-06, | |
| "loss": 1.5388, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "learning_rate": 6.079114751241613e-06, | |
| "loss": 1.5467, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "learning_rate": 6.024658011675526e-06, | |
| "loss": 1.5391, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "learning_rate": 5.970201272109437e-06, | |
| "loss": 1.5364, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "learning_rate": 5.915744532543348e-06, | |
| "loss": 1.5376, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 3.31, | |
| "learning_rate": 5.861287792977259e-06, | |
| "loss": 1.5397, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "learning_rate": 5.806831053411171e-06, | |
| "loss": 1.5336, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "learning_rate": 5.752374313845082e-06, | |
| "loss": 1.5378, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "learning_rate": 5.697917574278993e-06, | |
| "loss": 1.5318, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 3.49, | |
| "learning_rate": 5.643460834712905e-06, | |
| "loss": 1.5252, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 3.53, | |
| "learning_rate": 5.589004095146816e-06, | |
| "loss": 1.5333, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "learning_rate": 5.5345473555807275e-06, | |
| "loss": 1.5299, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 3.62, | |
| "learning_rate": 5.4800906160146385e-06, | |
| "loss": 1.5215, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 3.66, | |
| "learning_rate": 5.4256338764485495e-06, | |
| "loss": 1.52, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "learning_rate": 5.3711771368824605e-06, | |
| "loss": 1.5258, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "learning_rate": 5.316720397316373e-06, | |
| "loss": 1.5256, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "learning_rate": 5.262263657750284e-06, | |
| "loss": 1.5205, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "learning_rate": 5.207806918184195e-06, | |
| "loss": 1.5236, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 3.88, | |
| "learning_rate": 5.153350178618106e-06, | |
| "loss": 1.5281, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "learning_rate": 5.098893439052017e-06, | |
| "loss": 1.5175, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "learning_rate": 5.044436699485928e-06, | |
| "loss": 1.5215, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 1.5107132196426392, | |
| "eval_runtime": 3.2672, | |
| "eval_samples_per_second": 41.625, | |
| "eval_steps_per_second": 5.203, | |
| "step": 45908 | |
| }, | |
| { | |
| "epoch": 4.01, | |
| "learning_rate": 4.98997995991984e-06, | |
| "loss": 1.5202, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 4.05, | |
| "learning_rate": 4.935523220353751e-06, | |
| "loss": 1.5136, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "learning_rate": 4.881066480787663e-06, | |
| "loss": 1.5119, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 4.14, | |
| "learning_rate": 4.826609741221574e-06, | |
| "loss": 1.5052, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "learning_rate": 4.772153001655485e-06, | |
| "loss": 1.5088, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 4.23, | |
| "learning_rate": 4.717696262089397e-06, | |
| "loss": 1.5078, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "learning_rate": 4.663239522523308e-06, | |
| "loss": 1.5099, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 4.31, | |
| "learning_rate": 4.608782782957219e-06, | |
| "loss": 1.5098, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 4.36, | |
| "learning_rate": 4.5543260433911305e-06, | |
| "loss": 1.5044, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "learning_rate": 4.4998693038250415e-06, | |
| "loss": 1.5049, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 4.44, | |
| "learning_rate": 4.445412564258953e-06, | |
| "loss": 1.4958, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 4.49, | |
| "learning_rate": 4.390955824692864e-06, | |
| "loss": 1.5073, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 4.53, | |
| "learning_rate": 4.336499085126776e-06, | |
| "loss": 1.5015, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 4.57, | |
| "learning_rate": 4.282042345560687e-06, | |
| "loss": 1.5022, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 4.62, | |
| "learning_rate": 4.227585605994598e-06, | |
| "loss": 1.4973, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 4.66, | |
| "learning_rate": 4.17312886642851e-06, | |
| "loss": 1.4944, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 4.71, | |
| "learning_rate": 4.118672126862421e-06, | |
| "loss": 1.5024, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 4.75, | |
| "learning_rate": 4.064215387296332e-06, | |
| "loss": 1.4981, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 4.79, | |
| "learning_rate": 4.009758647730244e-06, | |
| "loss": 1.4942, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 4.84, | |
| "learning_rate": 3.955301908164155e-06, | |
| "loss": 1.4964, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 4.88, | |
| "learning_rate": 3.900845168598066e-06, | |
| "loss": 1.4995, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 4.92, | |
| "learning_rate": 3.846388429031978e-06, | |
| "loss": 1.4959, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 4.97, | |
| "learning_rate": 3.7919316894658886e-06, | |
| "loss": 1.4922, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 1.4904537200927734, | |
| "eval_runtime": 3.2628, | |
| "eval_samples_per_second": 41.682, | |
| "eval_steps_per_second": 5.21, | |
| "step": 57385 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 91816, | |
| "num_train_epochs": 8, | |
| "save_steps": 500, | |
| "total_flos": 8.99595804672e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |